The long-range goal of this work is to improve overall survival in hematopoietic stem cell transplant (HCT) recipients by personalizing their conditioning regimen and/or intravenous (IV) busulfan (BU) doses using metabolomics. BU is an essential part of the majority of HCT conditioning regimens, but has a narrow therapeutic index; low BU plasma exposure (caused by rapid clearance) is associated with an increased risk of rejection or relapse, while high BU plasma exposure is associated with an increased risk of hepatotoxicity. Although pharmacokinetic (PK)-based dosing to a target BU exposure is often conducted, relapse and toxicity continue to be problematic. Furthermore, shorter IV BU courses necessitate alternative methods to personalize IV BU. IV BU clearance, however, is not associated with polymorphisms of glutathione S-transferase (GST) A1 and GSTM1, the predominant GSTs in BU conjugation with glutathione. Therefore, we seek to test our working hypothesis that metabolomic signature, which provides novel insight into the in vivo cellular response and metabolite identification, is associated with IV BU clearance. We will first determine if endogenous metabolomics-based biomarkers obtained before BU administration can predict IV BU clearance. We will evaluate endogenous biomarkers using a targeted (glutathione pathway) and global analyses via liquid chromatography-mass spectroscopy. In addition, we seek to evaluate IV BU metabolism using metabolomics in parallel with gas chromatography-mass selective detection in patients receiving PK-based IV BU. We will evaluate plasma concentrations of eight different metabolites, building upon our experience with tetrahydrothiophene (THT+). We will also evaluate hydroxyTHT, which was just discovered by our group, and two recently identified metabolites that may contribute to BU toxicity: S-glutathione sulfonium conjugate (GS+THT) and ?-glutamyldehydroalanylglycine (EdAG). To complement this metabolomic work, we seek to validate covariates associated with IV BU PK, specifically age and body size, to mitigate a typical challenge for metabolomics research: the lack of understanding of potentially confounding factors. Our center has been a reference clinical laboratory for PK-based BU dosing since 1996 and thus, we have the largest database of IV BU PK. Using population pharmacokinetic (popPK) analysis, we will validate covariates for IV BU PK to guide future metabolomic studies. This popPK analysis can immediately improve patient care by using covariates to more accurately estimate an IV BU starting dose (i.e., before PK-based dosing) and by creating a limited sampling schedule to more efficiently use PK-based dosing. These complementary aims, which seek to identify novel metabolomics-based biomarkers, could overcome a critical barrier to HCT conditioning of balancing between response and toxicity. Based on our compelling preliminary data, we expect to identify an endogenous metabolomic signature that will influence the choice of an IV BU starting dose with the intention of improving overall survival for patients receiving IV BU-containing HCT regimens.